working on a new build - questions about phase

squamishdroc

I'm working on a small 2 way build and am just in the crossover design stage. My question is about the tweeter phase - is the constant switching above 3k normal? The crossover is 2nd order on the woof and 3rd on the tweet. I'm pretty happy with the response but wondering if the phase response needs some attention. The measurements were taken with the drivers in the box with OmniMic.

Kornbread

From 4-6k?  I thought this was just an artifact of the way 360* of phase is displayed in a 180 chart? 

dcibel

What does that phase response tell you? How are your interpreting the data to determine that there is a problem?

squamishdroc

dcibel said:

What does that phase response tell you? How are your interpreting the data to determine that there is a problem?

I'm not sure if there is a problem or not - I just wasnt sure if that is normal or not. I just don't remember seeing that much phase difference on my last few builds. So much to learn still😁

Wolf

It is not much phase variance at all. It just has to be plotted that way on a 2D graph. Phase should be plotted on a cylinder, where the upper and lower most extreme angles are actually the same point. +180 = -180 degrees.

6thplanet

Shouldn't the woofer and tweeter phases be closer together at the crossover frequency? I usually strive for them to be as close to overlapping as possible.

Wolf

They really don't diverge that abruptly above the xover, the wrapping to plot in 2D just makes it look that way.

dcibel

squamishdroc said:

dcibel said:

What does that phase response tell you? How are your interpreting the data to determine that there is a problem?

I'm not sure if there is a problem or not - I just wasnt sure if that is normal or not. I just don't remember seeing that much phase difference on my last few builds. So much to learn still😁

Look closely and you should see that the phase isn't rapping as the kids say, it's just hovering around the 180 degree point, the difference from the top of the phase plot and the bottom is only 1 degree you see...

In any case, I woudn't worry about that at all, I'd be more concerned in making sure I have the correct amount of BSC, hard to determine exactly with just a far field blended response. A document from Jeff B on how to splice near field and far field data for accurate information down low:

https://drive.google.com/open?id=0B-GpgK57lLMNOTljaGdrNl83WW8

And another on determining acoustic offset for good measure.

https://drive.google.com/open?id=0B-GpgK57lLMNMWYwajNHWjduaE0

 

squamishdroc

dcibel said:

squamishdroc said:

dcibel said:

What does that phase response tell you? How are your interpreting the data to determine that there is a problem?

I'm not sure if there is a problem or not - I just wasnt sure if that is normal or not. I just don't remember seeing that much phase difference on my last few builds. So much to learn still😁

Look closely and you should see that the phase isn't rapping as the kids say, it's just hovering around the 180 degree point, the difference from the top of the phase plot and the bottom is only 1 degree you see...

In any case, I woudn't worry about that at all, I'd be more concerned in making sure I have the correct amount of BSC, hard to determine exactly with just a far field blended response. A document from Jeff B on how to splice near field and far field data for accurate information down low:

https://drive.google.com/open?id=0B-GpgK57lLMNOTljaGdrNl83WW8

And another on determining acoustic offset for good measure.

https://drive.google.com/open?id=0B-GpgK57lLMNMWYwajNHWjduaE0

 

Thanks - I did the offset one and I will try the near/far one tonight if I get some time👍

4thtry

Edit:  I often get a lot of phase bouncing at higher frequencies on the woofer plot.  Tweeter plot is usually stable.  Sometimes I can clean up the phase bouncing by reversing the driver phase during testing.